Inside an ethical dilemmas case study: four weeks of not deciding

The firm below does not exist and none of this happened. It is an invented engineering consultancy, written so that a genuine conflict of duties can be followed without any real client attached.

What makes it worth reading is that nobody in it behaved badly. Every person acted on a real obligation, and the delay produced by their disagreement did more damage than any of the options would have.

What to take away

The situation

The consultancy did structural assessments. It had two clients in the same sector who did not know they shared an adviser, which was normal and covered by the engagement terms.

Working for the first client, a senior engineer found a fault pattern in a component type. It was not a failure but a way the component degraded under a loading condition absent from the original specification.

The finding applied only to work the first client paid for. It was commercially valuable to them because it let them plan replacements before failures occurred.

Six weeks later the same engineer was assigned to the second client, who was about to commit to a large purchase of the same component type for an application with that loading condition.

Why it was a genuine dilemma

The engineer could not tell the second client what he knew. The knowledge was the first client's, obtained at their expense, under terms the firm had signed. Telling would breach a duty the firm held and would be a straightforward wrong to a client who had done nothing except pay for good work.

He also could not say nothing with a clear conscience. The second client was about to spend a great deal on something he had reason to believe would degrade in their conditions, and they were relying on his firm's judgment, which is what they were paying for.

Both obligations were real. Neither was a preference dressed as a principle. This is the distinguishing feature of a dilemma and the reason the standard advice, escalate to someone senior, produces a senior person with the same problem.

Four weeks of not deciding

The discussions were serious and went in circles. The room kept relitigating which duty was more important rather than looking for options, and commercial reality made it worse.

The second client was the larger account, so arguments for telling them were suspected of being commercially motivated, and arguments for silence of protecting the smaller relationship. Both suspicions were unfair and both were raised.

Meanwhile the engineer was still on the second client's engagement, which meant he was actively advising them while withholding something material. Each week that passed made his position worse and made any eventual disclosure look more like a cover-up that had ended than a judgment that had been reached.

Nobody was assigned to own the question. Nobody set a date by which it would be answered. Those two omissions caused most of the damage.

The option nobody proposed

In week four a junior member of the team asked why they had not simply gone back to the first client and asked for permission to warn the second.

The room first assumed the first client would refuse. On review, nobody could explain why. The information gave no competitive edge over the second client, who was not a competitor.

The first client mainly wanted not to pay for research then given away. Asking the second client to commission the same work could solve that.

That is what happened. The first client agreed in a single phone call, on the condition that the second client pay for the analysis rather than receive it. The second client commissioned the work, changed the specification, and was told, correctly, that their adviser had spotted a risk and had arranged for the underlying analysis to be made available.

The fourth option honored both duties almost completely, and it had been available since the first week.

Why nobody saw it

Three reasons, all common.

The question was framed as a binary from the first conversation. "Do we tell them or not" is a question with two answers, and a room answering it will not generate a third.

The people arguing debated which duty ranked higher. That question is interesting, but it was not the one needing an answer. When two obligations conflict, the productive move is to find an option that reduces the conflict instead of resolving it by ranking.

The four lenses formalize this. Set out compactly in an introduction to ethical decision making, each generates different candidate options. That is most of their practical value.

And the junior person who saw it had been in the room from week two without being asked anything.

What it cost

The direct cost was four weeks of a senior engineer's discomfort and some partner time. The real costs were less visible.

The second client's purchase decision was delayed close to their deadline, which strained the relationship the firm had been protecting. The engineer, who had behaved correctly throughout, spent a month believing his firm might instruct him to stay silent, and told colleagues so. Two of them concluded that raising something like this was not worth the trouble.

And the firm learned nothing until someone wrote the case up afterwards, at which point the fourth option looked obvious to everyone who read it.

What changed

A named owner on the day. A conflict of duties is assigned to one partner within twenty-four hours of being recognized, with a date for an answer. Not to decide alone, but so that somebody's job is to reach an end.

Removal from the affected engagement. The person holding the conflicting knowledge comes off the second engagement while it is unresolved. This protects them and removes the compounding problem of advising while withholding.

A required third option. Any paper on a conflict of duties must state at least one option beyond the binary, and "ask the party whose permission would dissolve the conflict" is the first one to try.

Ask the most junior person present. Formally, and before the senior people speak.

The first and third are the ones that would have shortened this from four weeks to one. The general shapes these conflicts take, and what tends to break the tie in each, are set out in ethical dilemmas, and the reasoning structures behind them in ethical frameworks.

Where this case does not apply

If the first client's information had concerned wrongdoing rather than a technical finding, this would have been a different situation with different rules.

There may also have been legal protections for the person raising it, like those in the UK's whistleblowing framework for one jurisdiction. Confidentiality obligations and their limits are legally governed and differ by jurisdiction and profession. A firm facing a real version of this should talk to counsel in week one, not week four.

What the individual in the middle carries while all of that happens is the subject of employee ethics, and who is supposed to shorten it for them is in leadership ethics.

Common questions

Was the engineer wrong to stay silent for four weeks?

No. He raised it immediately and kept raising it. The failure was the firm's, for leaving him on the engagement without an owner or a deadline.

What if the first client had refused?

Then the firm would have faced the original binary, and would have had to decide, record the reasoning, and probably resign the second engagement rather than continue advising while withholding. Worse, but still better than a fourth week of nothing.

Is asking permission always available?

Not always, but it is available more often than rooms assume, and it costs one conversation to find out. The habit worth building is to test it before ranking the duties.